Eight products. One platform. One standard.

The flagship ambient scribe. A live session, dictation, or uploaded audio becomes a signed-ready note in your EHR’s own template, in seconds.

• Chart-aware by design. Longitudinal patient context informs every note, and the output lands in your organization’s own template structure rather than a generic form.
• Capture wherever the session happens. In person, telehealth, group, and offline on mobile. Native desktop capture for Windows and Mac removes the tab-and-screen-share dance that telehealth capture used to require.
• Group sessions, attributed per speaker. Group capture separates speakers so each individual’s note is scoped to that person’s own contribution — which is exactly what a group note is allowed to contain.
• Advanced explainability. Open any field and trace what the AI considered: the part of the transcript and the organizational context behind that specific line.

• Live session, dictation, or audio upload
• Native desktop capture for Windows and Mac, built for telehealth
• Group sessions with per-speaker attribution
• Trace any field back to the transcript and context behind it
• Discrete and narrative fields, tuned per discipline and program

Real-time and retrospective compliance. Point-of-submission checks that coach before a note becomes a denial, plus bulk audits across the chart base.

• Coverage no manual process can match. 100% chart coverage against roughly 10% for manual auditing, at five to ten seconds per check. Sampling cannot tell a leader what is happening across an enterprise.
• Two modes, one engine. Point-of-submission coaching for clinicians, retrospective bulk audit for compliance teams — both judged against the same organizational standard.
• Agentic re-write underway. The next generation pulls organizational DNA directly, with revenue-cycle capability folding in behind it.

• Save-time and submit-time checks
• Medical-necessity coaching in the EHR
• Your standards, enforced consistently

The work that lives outside the EHR: utilization-review packets, appeal letters, correspondence, and retrospective chart audits, assembled from the chart. Available today and included with the platform.

• Customers rank it above the scribe. A large California CCBHC called it more valuable than the scribe itself. Much of behavioral health’s administrative burden happens outside the EHR, and this is built for exactly that work.
• Upload anything, generate from the chart. OCR any source document, then produce templated output grounded in the record rather than in a blank form.
• Retrospective audits today, at scale next. Longitudinal chart audits work now, one document at a time. The version being built runs them in bulk across the historical documentation base and lands the output in the data warehouse, to power reporting the EHR was never built to do.
• Versatile enough that customers bend it. Organizations already build supervision templates and intake processes into it. It is the widest surface on the platform.

• Retrospective chart audits
• Medical-necessity review
• Denial appeal packets

The infrastructure layer. EHR vendors and health systems embed our models and orchestration inside their own products, white-label.

• Five EHR vendors already deploying. The embed motion is live and repeating, which makes it a distribution channel rather than a one-off integration.
• Same substrate underneath. Partners inherit the model stack, the configuration layer, and the governance layer — not a thinner version of the product.
• One source of clinical truth. No double entry between your systems and ours.

• SDK and APIs
• One source of clinical truth
• No double entry

Organization-specific clinical intelligence. Chat plus agentic patient and session lookup across your own records, inside your own governance.

• Grounded, cited answers. Responses carry citations rather than unsourced claims, so an answer can be checked instead of trusted blindly.
• Agentic lookup across your records. Queries reach the organization’s own patients and sessions, never outside them.
• The knowledge layer, not a chatbot. It is cross-cutting: it serves every product surface rather than sitting in one workflow. It runs in a side panel beside the EHR instead of over it, and pops out into its own window when you want to keep it open.

• Grounded in your org and session data
• Runs in a side panel beside the EHR, not over it
• Pop out into its own window and keep working
• Treatment-plan and approach support

Revenue integrity built on the documentation you already have: continuous eligibility, exemption proof, and appeals.

• Continuous eligibility monitoring. Coverage lapsing mid-treatment makes care unbillable. Eligibility is checked per patient on a configurable cadence, automatically or on demand, and every check persists to a workable dashboard. Eligibility is a billing function and a clearinghouse transaction, which is why it lives here rather than at intake.
• Exemption proof is the differentiator. Plenty of tools can check eligibility. Substantiating an exemption requires the clinical documentation behind it — which is what we already sit on.
• Denial and appeal workflows. Chart-grounded generation tied to the provider, with verification of benefits and medical-necessity alignment.

Scope is deliberate: eligibility, verification of benefits, medical necessity, and denials and appeals. Not prior authorization, claims, coding, or accounts receivable.

• Claim-readiness review
• Payer-rule alignment
• Denial prevention and rework triage

AI front office and intake. An AI receptionist that answers calls, plus referral processing that turns an inbound packet of records into a program and level-of-care recommendation.

• An AI receptionist, built in. Answers calls including after hours, sends forms, and monitors the intake inbox so referral workflows keep moving when no one is watching them.
• Referral processing and summarization. Referrals arrive at the intake department with a stack of medical records. Those get read and summarized rather than queued for someone to work through by hand.
• Program and level-of-care recommendations. Based on the clinical presentation in those records, matched against the programs your organization actually runs.

• Referral intake and triage
• Screening and eligibility capture
• Assessment drafted before the first session

Clinical supervision and oversight for the large unlicensed and pre-licensed behavioral-health workforce.

• Supervision, structured. Transcript aggregation, supervision notes, intervention alignment, and supervision-hour tracking in one place.
• Drift becomes visible early. Supervisors see where documentation and practice patterns diverge, with the evidence attached, so coaching happens on a cadence instead of after an audit.
• A category that barely exists. Purpose-built supervision tooling is close to absent in behavioral health, and the workforce that needs it most is the one still earning hours.

• Practice-pattern review
• Documentation coaching queues
• Evidence attached to every flag